This project will develop a self-contained portable testing system for the purpose of assessing individual performance in the workplace to detect impairment due to alcohol. The testing units will be based on programmable microcontroller technology, and will measure vigilance attention and reaction time performance. Impairment in these areas of function is likely to be the major reason for the increased risk of accidents from alcohol misuse and abuse in the workplace. Being able to efficiently detect such impairments has the potential to produce significant savings for society, both in economic and human costs. The economic costs have been estimated as high as $80 billion per year in the United States, so that even a small reduction in accident rate would be worthwhile. The project will take existing software and hardware components, currently operating on standard personal computers, and integrate these into a small, self-contained, portable testing unit. Task learning and convergence to baseline performance, day to day test-retest stability, and performance effects of ethanol will be studied in normal subjects, and a best possible multivariable discriminant function model will be developed to use as the testing system assessment algorithm. The key concept is comparison of individuals to their own baseline performance patterns in order to maximize accuracy of assessment. The assessment of individuals requires 2-3 minutes, provides immediate feedback, and is generally not considered an invasion of privacy, in contrast to urine testing, which is problematic in several respects.